We investigated the etiology of Leigh syndrome in 67 Australian cases from 56 pedigrees, 35 with a firm diagnosis and 32 with some atypical features. Biochemical or DNA defects were determined in both groups, ie, 80% in the tightly defined group and 41% in the "Leigh-like" group. Eleven patients had mitochondrial DNA point mutations (nucleotide [nt] 8993 T to G, nt 8993 T to C, or nt 8344 A to G) and 1 Leigh-like patient had a heteroplasmic deletion. Twenty-nine patients had enzyme defects, ie, 13 respiratory chain complex I, 9 complex IV, and 7 pyruvate dehydrogenase complex (PDHC). Complex I deficiency is more common than recognized previously. Six PDHC-deficient patients had mutations in the X-chromosomal gene encoding the E1alpha subunit of PDHC. Parental consanguinity suggested autosomal recessive inheritance in two complex IV-deficient sibships. We found no strong correlation between the clinical features and basic defects. An assumption of autosomal recessive inheritance (frequently made in the past) would have been wrong in nearly one-half (11 of 28 tightly defined and 18 of 41 total patients) of those in whom a cause was found. A specific defect must be identified if reliable genetic counseling is to be provided.

Discovering the molecular basis of mitochondrial respiratory chain disease is challenging given the large number of both mitochondrial and nuclear genes involved. We report a strategy of focused candidate gene prediction, high-throughput sequencing, and experimental validation to uncover the molecular basis of mitochondrial complex I (CI) disorders. We created five pools of DNA from a cohort of 103 patients and then performed deep sequencing of 103 candidate genes to spotlight 151 rare variants predicted to impact protein function. We used confirmatory experiments to establish genetic diagnoses in 22% of previously unsolved cases, and discovered that defects in NUBPL and FOXRED1 can cause CI deficiency. Our study illustrates how large-scale sequencing, coupled with functional prediction and experimental validation, can reveal novel disease-causing mutations in individual patients.

No abstract